Rwanda - Agricultural Services Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 16693 IMPLEMENTATION COMPLETION REPORT RWANDA AGRICULTURAL SERVICES PROJECT (Credit 2026-RW) June 6, 1997 Agriculture Group 2 Africa Region This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS US$ 1.00 equivalent in Rwanda Franc (RF) 1990 83 1991 125 1992 133 1993 144 1994 220 1995 290 1996 306 WEIGHTS AND MEASURES Metric System FISCAL YEAR OF BORROWER January 1 - December 31 Vice-President C. Madavo Country Director N.O. Tcheyan Technical Manager J. Baah-Dwomoh Task Team Leader M. Muhtar  FOR OFFICIAL USE ONLY ABBREVIATIONS AND ACRONYMS ASP Agricultural Services Project BEA Basic Extension Agent BNR Banque Nationale du Rwanda BP Banque Populaire (also UPB Union des Banques Populaires or Cooperative Savings Bank) CFRC-IWACU Centre de Formation et de Recherche en Cooperatives (Cooperative Research and Training Center and NGO) DCE Direction de Contr6le et Evaluation (Directorate for Control and Evaluation in the Ministry of Agriculture) FAO/CP Food and Agriculture Organization/Cooperative Program with the World Bank, Rome GOR Government of Rwanda IBRD International Bank for Reconstruction and Development IDA International Development Association IFAD International Fund for Agricultural Development (Rome) IRD Integrated Rural Development ISAR Institut des Sciences Agronomiques du Rwanda (Agricultural Research Institute of Rwanda KORA An NGO whose name means "Work" MINAGRI Ministere de l'Agriculture. de l'Elevage et des Forets (Ministry of Agriculture, Livestock and Forests) MINFIN Ministere des Finances (Ministry of Finance) NGC National Coordination Committee NGO Non-Governmental Organization NSC National Seed Committee OCIR-Caf6 Office des Cultures Industrielles du Rwanda Cafe (Office for Industrial Crops Rwanda - Coffee) OCIR The Office for Industrial Crops-Tea PCCV Projet Cafe et Cultures Vivrieres (Lake Kivu Coffee and Foodcrop Project) SMS Subject Matter Specialist SSS Service des Semences Selectionnees (Selected Seed Service) T&V Training and Visit Method of Extension UNDP United Nations Development Program UO Unite Operationnelle (Operational Unit) USAID United States Agency for International Development This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.   RWANDA AGRICULTURAL SERVICES PROJECT (Cr. 2026 RW) TABLE OF CONTENTS 1. PREFACE 2. EVALUATION SUMMARY A. INTRODUCTION ........................................................... I B. PROJECT OBJECTIVES .......................................................... C. IMPLEMENTATION EXPERIENCE AND RESULTS ........................................................... 11 D. SUMMARY OF FINDINGS, FUTURE OPERATIONS AND KEY LESSONS LEARNED ..........................1.................... 1ii 3. PART I: PROJECT IMPLEMENTATION ASSESSMENT A. STATEMENT AND EVALUATION OF OBJECTIVES ..........................................................1I Introduction ......................................................... I Project Objectives, Components, Covenants, Costs and Financing .........................................................1 B. ACHIEVEMENT OF PROJECT OBJECTIVES ...........................................................2 C. MAJOR FACTORS AFFECTING THE PROJECT .......................................................... 4 D. PROJECT SUSTAINABILITY ...........................................................6 E. BANK PERFORMANCE ...........................................................6 F. BORROWER ...........................................................7 G. ASSESSMENT OF OUTCOME ...........................................................7 H. FUTURE OPERATIONS .......................................................... 7 1. LESSONS LEARNED ...........................................................8 4. PART II: STATISTICAL ANNEXES TABLE 1: Summary of Assessments .9 TABLE 2: RELATED BANK CREDITS .11 TABLE 3: PROJECT TIMETABLE .II TABLE 4: CREDIT DISBURSEMENTS: CUMULATIVE ESTIMATED AND ACTUAL .12 TABLE 5: KEY INDICATORS FOR PROJECT IMPLEMENTATION .13 TABLE 6: KEY INDICATORS FOR PROJECT OPERATION .14 TABLE 7: STUDIES INCLUDED IN PROJECT .14 TABLE 8A: PROJECT COSTS .14 TABLE 8B: PROJECT FINANCING. 1 5 TABLE 9: ECONOMIC COSTS AND BENEFITS .15 TABLE 10: STATUS OF LEGAL COVENANTS .16 TABLE I 1: COMPLIANCE WITH OPERATIONAL MANUAL STATEMENTS .17 TABLE 12: BANK RESOURCES: STAFF INPUTS .17 TABLE 13: BANK RESOURCES -- MISSIONS .18   IMPLEMENTATION COMPLETION REPORT RWANDA AGRICULTURAL SERVICES PROJECT (Cr. 2026 RW) 1. PREFACE 1.1 This is the Implementation Completion Report for the Agricultural Services Project in Rwanda for which Credit 2026 RW in the amount of SDR 15.5 million (US$19.9 million equivalent) was approved on May 30, 1989 and made effective on June 15, 1990. 1.2 The credit was closed on December 31, 1996 which represented an extension from the original closing date of June 30, 1995. Final disbursement took place on 23/5/97, at which time, the cumulative balance of SDR 379 (US$527 equivalent), representing the undisbursed amount, was canceled. 1.3 A draft ICR was prepared by a FAO/World Bank Cooperative Program (FAO/CP) mission' which visited the country in August 1996. This draft was substantially revised and updated by Mansur Muhtar with support from Bhagwanti Khushalani and Isabelle Razafimanalina, based on additional information in project files as well as Country Team input. The borrower contributed to the preparation of the ICR by arranging field visits and meetings, providing data and information as well as comments on the mission's aide-memoire, The borrower also provided comments on the ICR which have been incorporated in the text. 1.4 The preparation of the ICR has suffered from insufficient availability of physical, economic and financial data as well as other information related to project performance, which was destroyed during the serious civil disturbances of 1994. Mr. F. Yriarte (Mission Leader) and Mr. H. Trupke (Marketing Economist).   IMPLEMENTATION COMPLETION REPORT RWANDA AGRICULTURAL SERVICES PROJECT (Cr. 2026 RW) 2. EVALUATION SUMMARY A. INTRODUCTION 2.1 The Agricultural Services Project (PSA) was formulated to help raise Rwanda's agricultural productivity and boost farm output. It was designed to consolidate, streamline and refocus efforts towards providing improved, cost-effective agricultural services for the country by replicating, on a broader geographic scale, the pilot experience under the IDA-financed Gitarama Agricultural Production project (PAG). Since 1973, the Bank has had a substantial involvement in the sector by providing support for multi-component area development projects. Due to the poor implementation experience with these types of projects, and the desire to institutionalize more coordinated approaches, the Bank's assistance strategy was being reoriented, by the late 1980s, towards supporting more focused stand-alone projects that would provide a common framework for donor assistance. The PSA belongs to this later generation. B. PROJECT OBJECTIVES 2.2 Project Objectives and Components. The main objectives of the project were to: (a) improve the delivery of extension services to farmers; (b) strengthen adaptive and on-farm research; (c) strengthen the production and distribution of selected seeds; and (d) promote greater private sector participation in the marketing of agricultural inputs and produce. The project replicated, on a broader geographic scale, the pilot experience under the IDA-financed Gitarama Agricultural Production project (PAG). To achieve its objectives, it was expected to finance, over a five-year period, the following components: (i) reorganization and streamlining of the extension services, including the implementation of a new structure based on the T&V extension principle; (ii) strengthening of adaptive and on-farm research; (iii) support for monitoring and evaluation; (iv) strengthening of seed production and distribution; and (v) promotion of private sector participation in input supply and marketing. Following the first supervision mission in 1991, the project's geographic scope was expanded to cover eight prefectures instead of the five originally planned. The proposed support for seed production and distribution was also reoriented away from direct public sector involvement to support for private participation, in line with the evolving role of Government within the context of a structural adjustment program supported by the Bank. 2.3 Evaluation of Project Objectives. Project objectives were broadly consistent with sectoral constraints/needs, and reflected the Government's stated agricultural development priorities. They were however too broad and ambitious to be undertaken within the framework of a single  ii intervention, given the country's weak absorptive and implementation capacity. Furthermore, while there was support for the project objectives from key political actors at the time, the voices of dissent were also significant, requiring more efforts in consensus building. The Gitarama pilot project upon which project design was based, was already exhibiting implementation and performance problems at a mid-term evaluation, which was conducted prior to the appraisal of the PSA project. The judgment was that these "teething" problems could be surmounted. In retrospect, the findings should have elicited a higher degree of caution in the design, scope and timing of the operation. In addition, the subsequent decision to expand the geographic coverage of the project, following a government request, was also inopportune. As originally agreed during project negotiations, this should have been preceded by a thorough evaluation at mid-term review. C. IMPLEMENTATION EXPERIENCE AND RESULTS 2.4 Achievement of Project Objectives and Sustainabilitx The overarching goal of improving agricultural productivity and boosting food supply was not achieved by the project. The Project- specific objective relating to improving institutional arrangements for agricultural services was only partially achieved. Progress made in strengthening seed production and promoting private sector participation in input and output marketing was either marginal or non-existent. At mid-term review, measures were identified for refocusing and redynamizing the project to achieve greater impact. These were however thwarted by the war of April 1994 and the subsequent which had tragic consequences on institutions, infrastructure, logistical and human capacity, as well as on social structures and relationships. Disbursements were resumed after the war, and project's closing date extended, to support sector rehabilitation and permit rebuilding of institutions. This decision was informed by the desire to provide rapid support to post-conflict reconstruction through existing vehicles, without the associated delay of preparing a new operation. As it turned out, the Bank underestimated the degree of effort required for jump-starting activities in the prevailing circumstances. Project activities are not sustainable in the present context, especially given the limited achievements recorded prior to the war-related dislocations. 2.5 Project Costs and Financing. At appraisal, total project costs were estimated at US$30.1 million, with a foreign exchange component of about 37 percent. The project was to be financed by an IDA credit of US$19.9 million, with expected domestic contribution amounting to US$10.2 million. As a result of implementation delays and war-related disruptions, as well as the substantial downscaling of infrastructure and private sector support programs, project-related expenditures amounted to only US$13 million, with an IDA financing of US$11.4 million. 2.6 Major Factors in Achievement of Objectives. The tragic events witnessed by Rwanda since 1990 have had the most profound impact on project implementation and limited the attainment of project objectives. A civil war broke out in October 1990, three months after credit effectiveness, which created an exceptionally difficult operating environment for the project and climaxed in the genocide of April 1994. The war and subsequent genocide of April 1994, together with ensuing events, resulted in massive human losses and associated social dislocations, decimation of institutional capacity, and destruction of physical investments, bringing project implementation to a complete halt. Attempts to restart project activities after the war proved futile, given the severely limited institutional capacity, fragile political balances and social dislocations. 2.7 The performance of the Government and the implementation agency also had a bearing on the achievement of project objectives. Problems encountered included: (i) delays and inadequacies in the provision of counterpart funds; (ii) wavering commitment on the part of the borrower;  iii (iii) interference of political factors in filling vacant key positions and associated conflictual relationships between political leadership and the technocrats: (iv) inability to make progress in liberalizing input supply and veterinary services, which hindereci the emergence of a competitive and service-oriented private sector; (v) administrative difficulties associated with insufficient understanding of roles and responsibilities; (vi) financial management improprieties; (vii) delays in submission of training and work plans; and (viii) insufficient attention to the systematic participation of the farmers. 2.8 Performance of the Bank and the Borzower. The PSA project was designed to replicate, on a broader scale, the pilot experience under the Gitarama Agricultural Services project, whose positive feature included the involvement of beneficiaries in project design and implementation. However, the Gitarama project suffered from poor institutional organization and bad management and its weak performance should have elicited a greater degree of caution in the timing of PSA preparation and appraisal, to allow for acquiring more experience. Instead, the Bank made the judgment to proceed with preparation on the optimistic assumption that the difficulties could be overcome in the follow- up project. Once it was decided to proceed, the preparation process was led by the borrower, based on extensive beneficiary consultation, which was innovative in the context. 2.9 Project implementation was constrained by weak borrower performance and inadequate covenant compliance. There were perpetual delays and shortfalls in providing counterpart funds, and Government commitment to the project appeared to be wavering at times. In addition, there was the persistence of an authoritarian culture, accompanied by weak beneficiary participation and a reluctance to support private sector development. There were also performance problems related to the selection and deployment of project staff, which was sometimes excessively influenced by political considerations. Amidst these constraints, the Bank made efforts to provide close implementation support from the field and was beginning to make some progress in introducing a structured, more disciplined approach to extension service delivery. The withdrawal of this support by the Bank, as well as the deteriorating security conditions, stymied these efforts. Mid-term review recommendations for addressing the performance problems could not be implemented. Following the war, the Bank and the Borrower restarted the project in a restructured form. However, they overestimated the capacity to move forward in rehabilitation institutions and eventually moved to close down the project. 2.10 Project Outcome. The project's outcome is rated as highly unsatisfactorv. Implementation was perpetually constrained by poor borrower performance, exacerbated by the difficult socio- political conditions prevailing in Rwanda. Only limited results were achieved by the mid-term review in 1993. Following the crisis of April 1994, the prospects for achieving project objectives and realizing development outcomes were jeopardized further. Even the modest institutional development gains realized under the project have been eroded. D. SUMMARY OF FINDINGS, FUTURE OPERATIONS AND KEY LESSONS LEARNED 2.11 Project Implementation Experience and Sustainability. Implementation continued to be hampered by fiscal, institutional and human resource constraints, exacerbated by wavering borrower commitment, persistence of an authoritarian culture, as well as administrative and political problems. The most significant factor shaping project implementation, however, was the endemic conflict that engulfed Rwanda since October 1990, and which culminated in the genocide of April 1994. In this context, it was impossible to continue with project implementation, which had, in any case, never  iv been easy under the protracted conflict conditions. Project sustainability, which was already questionable due to limited implementation progress realized, was jeopardized by these events. 2.12 Lessons for Future Projects. Any subsequent operation will have to take into account the changed parameters, while adequately incorporating lessons from the past. Some of the key lessons from this operation include: (a) project design should be kept simple, and tailored to the borrower's absorptive and implementation capacity; (b) borrower ownership and commitment is essential and should be continuously nurtured; (c) participation and empowerment of key stakeholders should be secured at various stages in the project cycle; (d) pilot operations, with clear benchmarks and indicators, should be used for improving knowledge about the operating environment and finding solutions to implementation problems, as well as for building stakeholder ownership, before mainstreaming based on sound professional judgment; and (e) the Bank should continuously assess operating risks and take appropriate action to hedge these in country situations characterized by adverse political and security conditions.  IMPLEMENTATION COMPLETION REPORT RWANDA AGRICULTURAL SERVICES PROJECT (Cr. 2026 RW) 3. PART I: PROJECT IMPLEMENTATION ASSESSMENT A. STATEMENT AND EVALUATION OF OBJECTIVES Introduction 3.1 The Agricultural Services Project (PSA) was formulated to help raise Rwanda's agricultural productivity and boost farm output. It was designed to consolidate, streamline and refocus efforts towards providing improved, cost-effective agricultural services, as a means of addressing the growing challenge posed by the country's excessive population pressure over limited arable land. With a population density of 400 per km2 of arable land (the highest in continental Africa), cropping practices, which centered around area expansion and increased cropping intensity, could no longer support sustainable agricultural growth. Annual agricultural output growth rate, which averaged 5.8 percent between 1974-80, had declined to 2 percent by the second half of the 1980s, against a population growth rate of 3.7 percent. Concomitantly, per capita daily kilo calories produced by the Rwanda farmers declined steadily from an already low level of 2,055 per person, reaching 1,509 per person by 1991. Against this background, it was considered necessary to boost productivity through the development and dissemination of relevant technology, complemented with other measures to improve inputs supply and marketing as well as expand credit access. The primary vehicle for achieving this goal was the PSA, which was appraised in 1988, building on IDA experience from previous and on-going operations in this sector. Project Objectives, Components, Covenants, Costs and Financing 3.2 j iect Objectives. Within the overall goal of government's strategy to increase productivity and output of the agricultural sector, the main objectives of the project were: (i) to improve the delivery of extension services to farmers with a particular emphasis on reaching women; (ii) to strengthen adaptive and on-farm research; (iii) to strengthen the production and distribution of selected seeds; and (iv) to promote greater private sector participation in the marketing of agricultural inputs and produce. The project was designed to replicate, on a broader geographic scale, the pilot experience under the IDA-financed Gitarama Agricultural Production project (PAG), whose objective was primarily to introduce an organizing framework for extension and adaptive research based on the Training and Visit extension principle. It was to be implemented over a five- year period (June 1990 to June 1995), with field-based activities covering selected sub-prefectures of Kigali, Kibuye, Kibungo, Gisenyi and Cyangugu. In 1991, following the first supervision mission, the  2 project's geographic scope was expanded to cover an additional three prefectures, leaving only two outside the project zone. Also, the planned support for public sector participation in input production and marketing was abandoned in favor of the private sector, largely as a result of the changes in the perception of Government's role ushered in by the structural adjustment process, and the subsequent progress in preparation of a proposed Agricultural Sector Adjustment Credit (AGSAC). 3.3 Project Components. To achieve its objectives, the project was expected to finance the following components: (i) reorganization and streamlining of the extension services, including the implementation of a new structure based on the T&V extension principles; (ii) strengthening of adaptive and on-farm research through better diagnosis of problems at the farm level and a participatory approach in problem identification and resolution; (iii) support for monitoring and evaluation through the implementation of a decentralized management system; (iv) strengthening of seed production and distribution; and (v) promotion of private sector participation in input supply and marketing through assistance to cooperatives and private enterprises. 3.4 Evaluation of Project Objectives. The objectives supported by the project appear broadly consistent with sectoral constraints/needs, and they reflect the Government's stated agricultural development priorities. There are a few shortcomings, however. First, these objectives may be considered too broad and ambitious to be undertaken within the framework of a single project intervention, given the country's weak absorptive and implementation capacity. Secondly, although project objectives and preparation process had the support of the political leadership at the time and involved extensive beneficiary participation, consensus within the Ministry of Agriculture was fragile. Thirdly, poor implementation results from the mid-term evaluation of the Gitarama pilot project, which was to provide the basis for the project design, should have elicited caution on the timing of project preparation. The prevailing judgment however was that these "teething" problems could be overcome through appropriate project design and intensive supervision efforts. Fourthly, the expansion of the geographic scope of the project further at an early stage during project implementation to support the establishment of a national extension system was premature. As originally agreed during project negotiations, this should have been preceded by a thorough evaluation at mid-term. 3.5 Project Costs and Financing. At appraisal, total project costs were estimated at US$30.1 million, with a foreign exchange component of about 37 percent. The project was to be financed by an IDA credit of US$19.9 million, with expected domestic contribution amounting to US$10.2 million. By March 1994, only US$8.5 million was disbursed against an original cumulative estimate of US$14.3 million. This disbursement lag was largely as a result of implementation delays and counterpart funding inadequacies, as well as the substantial downscaling of infrastructure and private sector support programs which were undertaken at the beginning of project implementation. The disruption of activities associated with the civil war and the tragic events of April 1994 also constituted a major explanatory factor. At the time of the closing of the credit account, US$11.4 million had been disbursed from the IDA Credit representing 52 percent of the SAR target. An amount of US$10.2 million was canceled, and the Government's total contribution amounted to about US$1.6 million, resulting in the re-estimated total project costs of US$13 million. B. ACHIEVEMENT OF PROJECT OBJECTIVES 3.6 The overarching goal of improving agricultural productivity and boosting food supply was not achieved by the project. Likewise, the 15 percent Economic Rate of Return estimated at appraisal is unlikely to be realized. As regards the specific objectives of the project, only partial  3 progress was made in improving the institutional arrangements for agricultural services delivery, while the strengthening of seed supply and distribution systems, as well as promoting private sector development were not achieved. In the absence of concrete data, it is difficult to quantify the extent of deviation from the stipulated objectives. But the following assessment of the implementation of various project activities supportive of the objectives illustrates the divergence between project objectives and outcome. 3.7 Reorganization and Streamlining of Extension Services. A major objective of the project was to help boost agricultural productivity through improved extension services. The focus was on the establishment of a nation-wide extension structure, based on the T&V principle, designed to streamline and harmonize public services in agriculture. There were initial start-up delays and problems associated with administrative and staffing difficulties, weak borrower commitment and misunderstanding of roles and relationships. With the resolution of these issues, intensified implementation support was provided by the Bank from an extension specialist located in the field. This began to have an impact in introducing a structured and disciplined approach to extension service delivery, especially at the provincial level, as well as nurturing ownership and fostering donor coordination. Some progress was also made in introducing the T&V principle on the ground. By early 1994, an estimated 20,000 demonstration plots had been established through about 6,000 farmers' contact groups, thus demonstrating the potential outreach of the system. 3.8 There were however continuing difficulties arising from the weak technology generation process and poor research linkages, which eventually resulted in rehashing of old techniques. Furthermore, diagnostic surveys with farmers planned to be carried out jointly with research, were not implemented, thus resulting in a disconnect between technical messages diffused by the extension agents and the concerns/interests of the farmers. There were also problems associated with shortages of qualified extension staff, low staff motivation and incessant counterpart funding problems. Other problems of systemic and policy nature included the persistence of a top-down, bureaucratic culture and weak community empowerment, and a lack of understanding and commitment to the system among some technical staff. The departure from the field of the implementation specialist and the deteriorating security situation further compounded the situation. The mid-term review, conducted in December 1993, proposed measures to address some of these problems, including: (i) the implementation of a staff training program; (ii) mainstreaming of participation. Its recommendations could not be implemented however, due to the turn of events in April 1994. The genocide and ensuing events further reduced the already insufficient number of extension staff available to 30 percent of the pre-war level, with many either killed or having fled the country. 3.9 Strengthening of Adaptive and On-farm Research. The Agricultural Research Institute of Rwanda (ISAR), in collaboration with the extension services, was to play an important role in diagnosing problems at farn level, implementing on-farm and multi-locational trials addressing farmers' priority problems, as well as conducting adaptive research. Improvements in the research- extension linkages were initially slow, as ISAR was involved in a reorganization process following the appraisal of the Second Agricultural Research Project. There were also delays in setting up Regional research centers, associated with the shortage of funds. Gradually, however, as these problems were overcome, more frequent meetings were beginning to result in useful collaboration, including the implementation of numerous scheduled on-farm trials. Unfortunately, time was too short for these results to concretize before the explosive events of 1994, which had devastating consequences on ISAR. Nearly all the researchers and technicians were killed, while the remaining fled the country, with only two presently remaining out of about sixty reconstituted staff after the  4 war. Moreover, stations were pillaged, buildings, vehicles and equipment were destroyed, and nearly all the germ plasma and scientific experiments lost. 3.10 Support for Monitoring and Evaluation. There were no plans for the creation of a separate monitoring and evaluation unit, although technical assistance from the UNDP was envisaged to strengthen the M&E function. The Monitoring and Evaluation system was ineffective, largely because the central management team, regional directorates, as well as operational units were subjected to frequent dismantling or reshuffling. There was also a lack of support by staff at all levels, who still continued to perceive the system as a means of policing and control. In these circumstances, the UNDP was unable to follow up on its initial action plan to strengthen the M&E services within the Ministry of Agriculture. Moreover, in the absence of pre-project studies to establish a base line of the main indicators for monitoring the project, it would have been difficult to monitor project outcome and impact. To address the deficiency above, a decision was taken in late 1993 to adapt the M&E system from an on-going UNDP/FAO project to PSA requirements, which would lead to the creation of a Monitoring and Evaluation Unit at the central level and in the regional offices. The design process for this proposal was however interrupted by the 1994 events. 3.11 Strengthening of Seed Production and Distribution. The availability of good quality seeds was seen as a key factor to improving agricultural productivity. Hence, the project sought to strengthen seed production and distribution as well as assist the implementation of a national seeds policy. Specifically, it sought to: (a) enhance the capacity of the Selected Seed Services (SSS) of MiNAGRI for improved seed quality control, certification and supervision; (b) facilitate the gradual rationalization of seed production through improved management of seed farms and the development of a multi-purpose farm that would serve as a focal point for small farmer contract outgrower schemes; and (c) provide support for reactivation of the National Seed Committee as a regulatory agency. At the first revision of the project (July 1991), it was agreed to phase out support for seed production by SSS and reorient actions towards facilitating the private sector, in accordance with the evolving perspective of the role of the state within the context of a structural adjustment program supported by the Bank. Accordingly, initiatives were taken in all Prefectures to form farmer groups and support their activities. Legal enactment of seed certification, however, was not carried out and input production continued to be heavily subsidized. 3.12 Promotion of Private Sector Participation in Input Supply and Marketing. The project aimed at strengthening the private sector provision of agricultural inputs as well as processing, storage and marketing of agricultural products. For that purpose, it was planned to: (i) create 5 agricultural promotion offices (BPA) initially acting as wholesalers; (ii) provide a guarantee fund to the "Banques Populaires" (BP) to facilitate the financing of the above mentioned activities, and (iii) train cooperatives and artisans. The BPAs were not created at the beginning and were abolished at the first revision. The credit operation with the BP never got started because of its lack of interest in financing agriculture. Training of cooperative managers and accountants as well as artisans began late (1992-93) and was reportedly registering successes, although no detailed information is available. This was however aborted with the war and subsequent departure of the Technical Assistance following the April 1994 events. C. MAJOR FACTORS AFFECTING THE PROJECT 3.13 Factors Not Generally Subject to Government Control. The tragic events witnessed by Rwanda in the last few years have had the most profound impact on project implementation and constitute a major contributory factor for non-attainment of objectives. A civil war broke out in  5 October 1990, three months after credit effectiveness, and the country was plunged into a spiraling cycle of violence and instability, which created an exceptionally difficult operating environment for the project. As schisms deepened further, the cycle of conflict and violence degenerated into full- scale genocide, which, together with ensuing events, resulted in massive human losses and associated decimation of public sector institutional/technical support capacity. Project implementation was brought to a complete halt and the basis for realizing its stipulated development objectives further eroded. 3.14 Following the war, the project closing date was extended to December 31, 1996 to help restore basic institutional capacity, rebuild infrastructure, as well as reconstitute logistical support needed for providing skeletal agricultural services. Slow progress was made however, given the changed parameters and the difficult country conditions. Specific challenges encountered include: (i) persistent counterpart funding shortages due to the erosion of the fiscal base; (ii) destruction of project investments; (iii) loss of vehicles, equipment and other logistical support; (iv) loss of technical and managerial staff, and associated loss of institutional memory; and (v) erosion of trust between farmers and technical staff. With the benefit of hindsight, the Bank, in its desire to respond to an emergency situation and provide quick support using existing vehicles, misjudged the complexity of the situation. 3.15 Factors Generally Subject to Government Control. The provision of adequate counterpart funds had been slow and erratic from the beginning of the project, resulting in delayed start-up. The problem was exacerbated by the outbreak of the war, with funding levels falling nominally as Government adjusted its budget to military requirements. The amendment of the Credit Agreement to increase the level of IDA financing did not overcome the problem, and project implementation continued to suffer delays. Government commitment to the project appeared to be wavering at the start, especially as the Gitarama project did not appear to be yielding the desired results. Because of this, there was an initial lack of interest in making available the required counterpart funds to make the project effective. The counterpart funding problem persisted through the project implementation period. The GOR was also slow in filling vacant key positions during various phases of the project, such as the position of the Director of Extension during the initial stages. In general, political factors ushered in by multi-partyism interfered in the selection and deployment of key technical staff, resulting in high staff mobility and creating discordant relationships. In the period following the genocide, protracted delays were experienced in filling the newly created position of project coordinator. Finally, the government failed to actively pursue its liberalization policy for input supply and veterinary services, which hindered the emergence of a competitive and service-oriented private sector. 3.16 Factors Generally Subject to Implementing Agency Control. During the initial stages, project implementation suffered from administrative difficulties associated with insufficient understanding of roles and responsibilities. In particular, the implementation of the extension services component was hampered by a lack of understanding on the part of the executing agency as to the philosophy and rules governing the T&V system, as well as a neglect of key organization and management features. The project manager suffered from work overload arising from his dual functions, and was often unable to secure the support of other directors at both regional and headquarters level. Financial management also suffered from improprieties on the part of the TA finance manager, who was subsequently dismissed. The M1NAGRI was slow in filling vacancies at headquarters, initiated staff training programs with substantial delays and was late in the submission of the project related work plan. Furthermore, the ministry paid little attention to the creation of an adequate M&E system, which could have facilitated timely identification of problems and  6 appropriate adjustments to project operation. Finally, community empowerment and farner participation were weak. Mid-term review recommendations aimed at correcting these deficiencies could not be implemented. D. PROJECT SUSTAINABILITY 3.17 Project sustainability is unlikely because of the limited progress made in implementing proposed project activities. As indicated earlier, only partial progress was made in institution building with regard to the introduction of a structured, disciplined extension system that was beginning to gain the support of other donors. Even in the limited areas where progress has been realized, future sustainability has been undermined by the deleterious consequences of the war on institutions, infrastructure, logistical and human capacity, social structures and relationships. Although attempts have been made to rehabilitate basic infrastructure and reconstitute the logistical base, this is by no means sufficient. Poor implementation progress of the other project components at the time of project closing, and the overall operating context associated with the impact of the conflict, render them unlikely to be sustainable. E. BANK PERFORMANCE 3.18 The PSA project was designed to replicate on a broader scale, the Gitarama Agricultural Services project (Cr. 1669 RW) which was prepared as a vehicle to test a new approach to the provision of agricultural services involving extensive beneficiary participation. The mid-term evaluation of the Gitarama project, which had just been completed at the time of PSA preparation, revealed implementation difficulties and problems. Nonetheless, the Bank proceeded with PSA project preparation and appraisal, on the assumption that the these problems could be surmountable with proper project design and intensive supervision effort. Accordingly, project preparation was done in a participatory manner. Implementation however proved difficult and was continuously hampered by similar problems identified in the implementation of the Gitarama project (as well as the previous projects), including scarcity of technical messages; weak extension-research linkages; insufficient beneficiary participation; project complexity; inadequate monitoring system. With the benefit of hindsight, it would have been better to delay PSA appraisal (or, at least, to limit the scope of the project) until there was a better understanding of why the Gitarama project was experiencing problems. 3.19 During the project's first supervision mission, amidst various implementation problems, the Bank acceded to the borrower's request to extend the geographic scope of the project to provide support for implementation of a national extension policy. Given implementation problems, the decision to extend the geographic scope also should have been delayed until a mid-term evaluation was carried out, as agreed during project negotiations. To assist the borrower in improving implementation performance, an extension specialist had been posted to the field by the Bank, and with time, noticeable results were being registered (i.e. demonstration plots and farmer contact groups). However, the withdrawal of this support and the deteriorating security conditions stymied these efforts. The Bank's supervision efforts with respect to Monitoring and Evaluation were deficient. Following the war, the Bank misjudged the capacity of the borrower to resume project implementation and hence did not adequately consider the option of closing the project. In its effort to provide immediate assistance to resumption of agricultural activity, the Bank agreed with the Borrower on a new action plan for restarting the project after the war. However, capacity had been  7 weakened to such an extent that this action plan could not be implemented. The Bank moved to close the project when it became obvious that results could not be achieved in this context. F. BORROWER 3.20 The Borrower's performance during project implementation was generally weak and covenant compliance inadequate. Project start-up was delayed because of the failure of Government to meet counterpart funding requirements (a major credit covenant), which remained a perennial problem throughout the implementation period. Government commitment to the project also appeared to be wavering at times, which exacerbated the problem. In addition, there were performance problems related to the selection and deployment of project staff, which were sometimes overly influenced by political considerations. Project management problems resulted in three qualified and/or unsatisfactory audits while the non-establishment of a proper M&E system deprived management of an effective tool for monitoring performance and adjusting project operations. The achievement of results on the ground was undermined by the persistence of an authoritarian culture, accompanied by weak beneficiary participation and reinforced by an occasional aversion to private sector involvement. The legal and regulatory reforms and enabling measures needed to promote privatization of veterinary services and input supply were not sufficiently addressed. The weak performance of the Borrower, however, has to be set against the background of the prevailing socio-political environment which made project implementation increasingly difficult from 1990 onwards. G. ASSESSMENT OF OUTCOME 3.21 The project suffered from slow start-up associated with major implementation problems. With intensified supervision efforts, some progress was being made in the reorganization of the extension service, although there were continuing difficulties with regard to weak extension-research linkages, insufficient beneficiary participation, staff shortages and recurrent funding problems. The departure of the extension specialist and the deteriorating security conditions in early 1993 eroded the gains that were being recorded. Little progress was made in strengthening seed production and promoting private sector participation in input and output marketing, while cost-recovery and subsidy reduction efforts were only half-heartedly supported. With the crisis of April 1994 and its aftermath, the prospects for achieving project objectives and realizing development outcomes were jeopardized further. As a result, anticipated project benefits including institutional development, increased productivity and output, financial gains, improved nutrition and poverty alleviation were not realized. There is considerable uncertainty over the sustainability of even the modest gains realized under the project. For these reasons, therefore, the project's outcome is rated as highly unsatisfactory. H. FUTURE OPERATIONS 3.22 There is a strong commitment by the GOR to provide continuing support to the agricultural sector. Discussions have been initiated with the Bank to support a follow-up operation. Given the complex, high risk operating environment and limited implementation capacity, a modulated approach, combining a blend of instruments, is being considered. One possibility is to initiate a pilot operation focusing on selected strategic areas such as agricultural services and community-based rural investments. The operation will be complemented by ESW and grant-funded capacity- consensus building activities to clarify and improve the framework for intervention in the sector.  8 Based on knowledge and experience acquired, a replication of efforts may be envisaged on a phased, calibrated manner, using repeater loans, or scaling up to major investments depending on the triggers to be built into the proposed project on policy making and capacity rebuilding. I. LESSONS LEARNED 3.23 Some of the key lessons from this operation include: (a) Project design should be kept simple, and tailored to the borrower's absorptive and implementation capacity. In this regard, preparation should be accompanied by intensive assessment of the fiscal and institutional context. (b) Borrower ownership and commitment is essential and should be continuously nurtured. (c) Participation and empowerment of key stakeholders should be secured at various stages in the project cycle - design, preparation, implementation and evaluation. (d) Pilot operations provide a potentially fruitful avenue for improving knowledge about the operating environment and finding solutions to implementation problems, as well as building ownership, commitment and familiarity with project concept among the stakeholders. Sufficient time should however be allowed to elapse for lessons and experiences to crystallize and for learning and internalization to take place. Clear benchmarks and indicators should guide the transition from pilots to full-scale operations, which underscore the importance of establishing functioning monitoring and evaluation systems. (e) Social conflict, adverse political and security conditions, have an overriding impact on project outcomes. While it is difficult to foresee these, Bank should adequately assess operating risks and take appropriate action to hedge against these.  9 IMPLEMENTATION COMPLETION REPORT RWANDA AGRICULTURAL SERVICES PROJECT (Cr. 2026 RW) 4. PART II: STATISTICAL ANNEXES TABLE 1: SUMMARY OF ASSESSMENTS A. Achievement of Objectives Substantial Partial Negligible Notapplicable (/) (/) ~~~ ~~(V) (VO Macro Policies 0 0 0 0 Sector Policies 0 0 El 0 Financial Objectives 0 0 El 0 Institutional Development 0 0 0 0 Physical Objectives 0 0 0 0 Poverty Reduction 0 C E0 0 Gender Issues 0 O 0 0 Other Social Objectives 0 0 0 0 Environmental Objectives 0 0 El 0 Public Sector Management 0 0 0 0 Private Sector Development 0 0 0 0 Trade and tariff reform 0 0 0 0  10 TABLE 1: SUMMARY OF ASSESSMENTS (Continued) B. Project Sustainability I iUnlikely Uncertin _______________ (v') ) HWI C. Bank Performwanc satisfato Satisaa Deficien Identification O 0 0 Preparation Assistance O0 0 0 Appraisal OE 0 0 Supervision El 0 21 Highly D. Borrower Perfomance satisfactrQy Satisfactor Deficie Preparation El 0 0 Implementation 0 0 0 Covenant Compliance __ 0 0 Operation (if applicable) 0 O 0 ighly E. Assessment of Outcome satisfactory Satisfactory Unsatisfactory unsatisfactory (V) (1) (1) (V) 0 El 0 0  11 TABLE 2: RELATED BANK CREDITS Loan/Credit Title Purpose Year of Status Approval Preceding Operations _ _ _ _ _ 1. Rural Development I, Area Development 1979 Completed 1986 (Cr.937 RW) 2. Integrated Forestry Project (Cr. Forestry 1980 Completed 1987 1039) Development 3. Lake Kivu Coffee and Foodcrops Support to Coffee, 1981 Completed 1988 Production Project (Cr. 1126 RW) and Food Crops Sector 4. Bugesera-Gisaka-Migango (BGM Area Development 1982 Completed 1991 11) (Cr. 1283 RW) 5. Agricultural Research Project I Support to 1985 Completed 1991 (Cr. 1546 RW) Agricultural Research 6. Gitarama Agricultural Services Support to 1986 Completed 1991 (Cr. 1669 RW) Agricultural Services Development 7. Forestry II (Cr. 1811 RW) Forestry 1987 Completed 1994 Development 8. Agricultural Research II Support to National 1993 Canceled 1994 (Cr. 2547 RW) Agricultural Research __ _ TABLE 3: PROJECT TIMETABLE Steps in Project Cycle Date Planned Date Actual/ Latest Estimate Identification (Executive Project Summary) March 30, 1988 Preparation/Pre-appraisal July 1988 Appraisal September 1988 October 1988 Negotiations April 1989 April21, 1989 Board presentation May 1989 May30, 1989 Signing September 13, 1989 Effectiveness December 1989 June 15, 1990 Midterm review (if applicable) December 1993 December 10, 1993 Project completion December 31, 1994 Not completed Loan closing June 30, 1995 December 31, 1996  12 TABLE 4: CREDIT DISBURSEMENTS: CUMULATIVE ESTIMATED AND ACTUAL (US$ thousands) ___________._______________ FY90 FY91 FY92 FY93 FY94 FY95 FY96 FY97' Appraisal estimated 0.4 2.3 6.2 11.0 15.3 19.1 19.9 19.9 Actual 0 2.2 2.8 6.2 8.5 8.5 9.8 11.4 Actual as percent of 0 96 45 56 56 45 49 52 estimated I I I _ __ Date of final May 23, 1997 disbursement 2 According to SAR no disbursements were foreseen in FY 1997. As a result of extension of the credit closing date and Bank's approval of selected expenditure beyond the revised date additional claims amounting to about US$ 1.3 million are expected to be submitted. 3 Includes US$ 1.0 million opening fund, actual cumulative disbursements up to end of FY 91 amounted to only US$ 1.2 million equivalent to 52 percent of target.  13 TABLE 5: KEY INDICATORS FOR PROJECT IMPLEMENTATION Indicators Unit Estimated Actual I. Key Implementation indicators in SAR/President's Report 1. Prefectures to be covered 5 8 2. Extension staff a) high level 113 b) lower level 248 3. Extension staff training a) local 54,600 m/days n.a b) overseas 990 m/days 4. Extension Service Vehicles a) cars/pick-ups 31 47 4 b) motorcycles 330 200 c) bicycles __ _ 143 1000 5. Forest sector nurseries 239 n.a 6. TA + consultants a) R/D 8 m/m 0 b) M/E 9 m/m I m/m c) training 13 rn/m 13 m/m d) finance 36 rn/m 36 m/m e) seed multiplication 5 9 m/m 9 m/m I. Modified Indicators _ 1. Prefecture to be covered 9 8 2. Extension staff a) high level 259 302 b) lower level 937 678 3. Extension staff training Allocation increased without specification of rn/days n.a All other quantifiable actions pertaining to community support, adaptive research, M&E, seed production, promotion of private initiative, women support program and animal productions were not only marginally tackled by the project and are therefore not listed here. The only notable exception was the training of artisans by an NGO under the private sector support initiative. After extension of geographical coverage. 5 Funded by CIDA.  14 TABLE 6: KEY INDICATORS FOR PROJECT OPERATION Not applicable TABLE 7: STUDIES INCLUDED IN PROJECT PIrpose J Status j Impact Marketing of Improvement of marketing Completed No impact as study was completed Agricultural system only after the events of April 1994 Products Processing of To identify scope for rural Not Agricultural agroprocessing Implemented Products Introduction of fruit To diversity agricultural ISAR6 and vegetables production base TABLE 8A: PROJECT COSTS (in US$ million) Appraisal estimate Actial estimate Item Local Foreign Total Local Foreign Total I. Investment Costs A. Construction 3.3 2.2 5.5 0.5 0.3 0.8 B. Equipment, Vehicles, 1.7 2.2 3.9 2.8 3.6 6.4 Materials etc. C. Consultants, Training and 0.9 2.6 3.5 0.5 1.5 2.1 Studies D. Promotion 2.0 0.5 2.5 0.0 0.0 0.0 E. Nursery Forest Cafe 2.3 0.6 2.8 0.2 0.0 0.2 Sub-total 10.2 8.1 18.3 4.0 5.5 9.5 II. Operating Costs 8.8 3.1 11.9 2.5 0.9 3.4 Total Project Costs 18.9 11.2 30.1 6.5 6.4 12.9 6 The study was initiated by ISAR under the Agricultural Research II Project ( Cr. 2547) but not yet completed.  15 TABLE 8B: PROJECT FINANCING (in US$ million) i_____________________ Appraisal estimate Actual estimate Source Local Foreign Total Local Foreign Total IDA 9.9 10.0 19.9 5.7 5.6 11.3 Government 9.0 1.2 10.2 0.8 0.8 1.6 Total 18.9 11.2 30.1 6.5 6.4 12.9 TABLE 9: ECONOMIC COSTS AND BENEFITS On assumption that annually 4 percent of families would adopt improved farming practices, it was estimated at appraisal that the project would achieve an ERR of 15 percent. Due to the interruption caused by the war and the subsequent near collapse of the extension system no attempt has been made to re-estimate the ERR, especially since no agricultural statistics covering the project period were available. Includes US$ 1.3 million in Taxes. 8 Includes US$ 70,000 in taxes (mission estimate)  16 TABLE 10: STATUS OF LEGAL COVENANTS Agreement Section Covenant Present Original fulfilment date Revised fulfilment date Description of Comments type status covenant DCA 2.02 b) I C Special Account 3.02 b) 9 CD Nov. 30 annually Annual Work Plan Repeatedly late 3.02 c) 9 C Nov. 30 1992 Nov. 30, 1993 Mid-tern Review Postponed by one year 3.02 d) 9 C July 31, annually Reporting 3.03 a) 4 CP January 1, annually Governunent funding Late and insufficient 3.03 b) (i) 4 C January 1, annually Project account 3.03 b) (i), (ii) 4 CD January 1, annually . Fund availability Repeatedly late 3.03 c) 3 CD January 1, annually Funding of local subaccounts Repeatedly late 3.04 10 C June 30, 1990 Extension, contracting to local communes l 3.05 10 NC Dec. 31, 1994 Forest nurseries, divestment Marginal progress made 3.06 a) and b) 2 NC Rental of housing to staff Staff resistance 3.06 c) and d) 2 C Leasing equipment to staff 3.06 e) 10 C Dec. 15, 1990 Gishwati Centre, privatisation proposals 3.07 10 CP Dec. 31, 1992 Seeds production, cost recovery, seed certification Process was under review prior to April 1994 3.08 (i) 10 CP March 31, 1990 Contracting credit to Banques Populaires and co- Limited loans extended but no records operative available 3.08 (ii) 10 CP March 31, 1990 Support activities to IWACU Limited training given 3.09 9 NC Dec. 31,1991 Monitoring and evaluation Late attempts to establish a system were swatted by the events of 1994 4.01 a) I CP Annually Maintain records and accounts Disrupted by events of April 1994 4.01 b) I CP June 30, annually Audit accounts, including Special Account, no later Disrupted by events of April 1994 than six months after end of fiscal year 4.01 c) I CP Annually Maintain and audit records of statements of expenditure Disrupted by events of April 1994 for which withdrawals have been made from the Credit Covenant types: I = Accounts/audits; 2 Financial performance/revenue generation from beneficiaries; 3 Flow and utilisation of project funds; 4 = Counterpart funding; 5 = Management aspects of the project or executing agency; 6 = Environmental covenants; 7 = Involuntary resettlement; 8 = Indigenous people; 9 = Monitoring, review, and reporting; 10 = Project implementation not covered by categories 1-9 Present status: C = covenant complied with; CD =complied with after delay; CP = complied with partially; NC =not complied with  17 TABLE 11: COMPLIANCE WITH OPERATIONAL MANUAL STATEMENTS No significant lack of compliance with applicable Bank Operational Manual Statements could be observed. TABLE 12: BANK RESOURCES: STAFF INPUTS Stage ofroject Cycle ACTUAL ......._________, _.. __. W eeks US$('000} Preparation to appraisal 61.7 122.6 Appraisal 32.1 82.2 Negotiations through Board-Effectiveness 10.6 27.3 Supervision 137.2 400.9 Completion 14.8 56.2 TOTAL 256.5 689.2  18 TABLE 13: BANK RESOURCES - MISSIONS Performance Rating Number Specialized Implementation Development Stage of Month/ of Days in Staff Skills Status Objectives Types of Project Cycle Year Persons Field Represented Problems (IDA) Identification 12/86 4 10 E; Al AE; FA. . Preparation Assistance 11/87 2 5 E; A Preparation Assistance 03/88 3 9 El A. FA Pre-appraisal 07/88 7 16 E. A. FA, ExS, SS; ME; RD Appraisal 10/88 3 12 E; A; FA Board approval through effectiveness Supervision 1 7/91 3 10 A; ExS; AF 3 3 Financial. Management, Legal Con.., Supervision II 5/92 3 5 A, FA; ExS 2 I Financial, Legal Cov. Supervision III 6/93 1 2 AE 2 2 Financial Supervision IV 12/93 5 19 FA; ExS; A; PS 2 2 All (Mid-Termn Review) Supervision V 6/95 3 15 AE; OA; A HU HtJ Financial, Management, Etc. Supervision VI 11/95 3 9 AE; OA: A HU HU Financial, Maniagement, Etc. Supervision VII 6/96 2 16 FA; A HU HU All Completion9 8/96 2 12 A; AE U Hl] E = Economist ME = Monitoring Evaluation A = Agriculturist RD Research Development AE = Agricultural Economist PS Privatization Specialist FA = Financial Analyst OA Operations Analyst ExS = Extension Specialist HU Highly Unsatisfactory SS = Seed Specialist U Unsatisfactory. Draft preparation by FAO/CP.  NOTES    IMAGING Report No.: 16693 Type: ICR